Adaptation of Genetic Algorithm to Determine Air Jet Shape of Flow Visualized by Helium Bubbles

During the development of ventilation and air conditioning systems, it is important to verify the functionality of the devised system, especially the method by which air fills the space. In this case, an optical visualization method of introducing helium bubbles into the airflow can be used for this function. A special generator produces small bubbles filled with helium. These bubbles are visible particles of airflow, making it possible to determine the shape of such a flow. This article describes the evaluation of flow images visualized by helium bubbles. Bubbles are detected by comparing intensities and subsequently cleaning faulty bubbles. For a convex airflow shape, exact geometric methods are used, and for a more precisely nonconvex shape of airflow, a genetic algorithm is used, in which a sophisticated algorithm of arranging a boundary line from selected points is applied. The algorithm also includes a specific approach of representing individuals in population inclusive of determining their fitness values.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01611274
  • Record Type: Publication
  • Files: TRIS, ASCE
  • Created Date: Jul 13 2016 3:02PM